MedVerse: Efficient and Reliable Medical Reasoning via DAG-Structured Parallel Execution

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Chen, Jianwen, Yang, Xinyu, Xia, Peng, Azarang, Arian, Lee, Yueh Z, Li, Gang, Zhu, Hongtu, Li, Yun, Chen, Beidi, Yao, Huaxiu
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915939141812224
author Chen, Jianwen
Yang, Xinyu
Xia, Peng
Azarang, Arian
Lee, Yueh Z
Li, Gang
Zhu, Hongtu
Li, Yun
Chen, Beidi
Yao, Huaxiu
author_facet Chen, Jianwen
Yang, Xinyu
Xia, Peng
Azarang, Arian
Lee, Yueh Z
Li, Gang
Zhu, Hongtu
Li, Yun
Chen, Beidi
Yao, Huaxiu
contents Large language models (LLMs) have demonstrated strong performance and rapid progress in a wide range of medical reasoning tasks. However, their sequential autoregressive decoding forces inherently parallel clinical reasoning, such as differential diagnosis, into a single linear reasoning path, limiting both efficiency and reliability for complex medical problems. To address this, we propose MedVerse, a reasoning framework for complex medical inference that reformulates medical reasoning as a parallelizable directed acyclic graph (DAG) process based on Petri net theory. The framework adopts a full-stack design across data, model architecture, and system execution. For data creation, we introduce the MedVerse Curator, an automated pipeline that synthesizes knowledge-grounded medical reasoning paths and transforms them into Petri net-structured representations. At the architectural level, we propose a topology-aware attention mechanism with adaptive position indices that supports parallel reasoning while preserving logical consistency. Systematically, we develop a customized inference engine that supports parallel execution without additional overhead. Empirical evaluations show that MedVerse improves strong general-purpose LLMs by up to 8.9%. Compared to specialized medical LLMs, MedVerse achieves comparable performance while delivering a 1.3x reduction in inference latency and a 1.7x increase in generation throughput, enabled by its parallel decoding capability. Code is available at https://github.com/aiming-lab/MedVerse.
format Preprint
id arxiv_https___arxiv_org_abs_2602_07529
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle MedVerse: Efficient and Reliable Medical Reasoning via DAG-Structured Parallel Execution
Chen, Jianwen
Yang, Xinyu
Xia, Peng
Azarang, Arian
Lee, Yueh Z
Li, Gang
Zhu, Hongtu
Li, Yun
Chen, Beidi
Yao, Huaxiu
Machine Learning
Large language models (LLMs) have demonstrated strong performance and rapid progress in a wide range of medical reasoning tasks. However, their sequential autoregressive decoding forces inherently parallel clinical reasoning, such as differential diagnosis, into a single linear reasoning path, limiting both efficiency and reliability for complex medical problems. To address this, we propose MedVerse, a reasoning framework for complex medical inference that reformulates medical reasoning as a parallelizable directed acyclic graph (DAG) process based on Petri net theory. The framework adopts a full-stack design across data, model architecture, and system execution. For data creation, we introduce the MedVerse Curator, an automated pipeline that synthesizes knowledge-grounded medical reasoning paths and transforms them into Petri net-structured representations. At the architectural level, we propose a topology-aware attention mechanism with adaptive position indices that supports parallel reasoning while preserving logical consistency. Systematically, we develop a customized inference engine that supports parallel execution without additional overhead. Empirical evaluations show that MedVerse improves strong general-purpose LLMs by up to 8.9%. Compared to specialized medical LLMs, MedVerse achieves comparable performance while delivering a 1.3x reduction in inference latency and a 1.7x increase in generation throughput, enabled by its parallel decoding capability. Code is available at https://github.com/aiming-lab/MedVerse.
title MedVerse: Efficient and Reliable Medical Reasoning via DAG-Structured Parallel Execution
topic Machine Learning
url https://arxiv.org/abs/2602.07529